Home / Repository / PRWP / RR_USA_2025_499
PRWP

Reproducibility package for How Do Organizations Learn? The Diffusion Of Scientific Evidence On Generative AI

2026
Get Reproducibility Package
Reference ID
RR_USA_2025_499
Author(s)
Mahvish Shaukat, Andreas Stegmann, Mattie Toma
Collections
World Bank Policy Research Working Papers
Metadata
JSON
Created on
Feb 04, 2026
Last modified
Feb 04, 2026
  • Overview
  • Reproducibility Package
  • Description
  • Scope and coverage
  • Disclaimer
  • Access and rights
  • Contacts
  • Information on metadata

    • Overview

    Abstract

    We study information diffusion in a large organization through a field experiment at the World Bank. We focus on transmission of scientific evidence on the impacts of generative AI by experimentally varying whether research findings are shared with senior or junior staff, and varying beliefs about peer adoption and evidence credibility. Providing evidence to senior staff significantly increases transmission and diffusion as measured by engagement with study materials and colleagues' recall of study details. In contrast, changing beliefs about peer adoption or credibility has no detectable effects. The results highlight the importance of organizational hierarchy in shaping informal information flows.

    Reproducibility Package

    Scripts
    Get Reproducibility Package
    File name
    RR_USA_2025_499
    Zip package
    RR_USA_2025_499.zip
    Title
    Reproducibility package for How Do Organizations Learn? The Diffusion Of Scientific Evidence On Generative Ai
    Date
    2026-01
    Dependencies
    Stata dependencies are listed in the ado folder.
    Instructions
    See README in reproducibility package.
    Notes
    Computational reproducibility verified by Development Impact (DECDI) Analytics team, World Bank.
    Source code repository
    Repository name URI
    Reproducible Research Repository (World Bank) https://reproducibility.worldbank.org
    Software
    Name
    Stata
    Version
    18 MP

    Reproducibility

    Technology environment

    OS: MacBook-Pro (13 inch, M2, 2022)
    Processor: Apple M2
    Memory available: 8 GB

    Technology requirements

    ~1 minute runtime.
    Reproduction instructions

    To reproduce the findings in this study, please follow the steps below:

    1. Acquire the Data. Please refer to the Datasets subsection of this documentation.
    2. Change main directory. Update the global macro $mainFolder to the root directory where your data and output folders reside.
    3. Run the master do file.

    Data

    Datasets
    Name
    How Do Organizations Learn Survey Data
    Note
    Dataset names: `Wave 1 Seed.dta`, `Wave 2 Seed.dta`, `Wave 1 Contact.dta`, `Wave 2 Contact.dta`, `Wave 1 Treatment.dta`, `Wave 2 Treatment.dta`. `Wave1_article_clicks.csv`, `wave1_infographic clicks.csv`, `wave2_article clicks.csv`, `wave2_infographic clicks.csv`, `Wave 1 Seed_batch.dta`, `Wave 2 Seed_batch.dta`, `Wave 1 Links_withNewID.dta`, `Wave 2 Links_withNewID.dta`, `Evidence Transmission_SSPP.xls`, `Evidence Transmission_WP.xls`. These raw data files contain personally identifiable information and are only available to the authors. An anonymized version of the dataset compatible with this reproducibility package will be made available at the World Bank Microdata Library in the future For Wave1_URL_links and Wave2_URL_links datasets users can reach out to mshaukat@worldbank.org. Access date: 2025.
    Access policy
    Data is forthcoming on the World Bank Microdata Library.
    Citation
    Shaukat, M., Stegmann, A., Tome, M. (2025). "How Do Organizations Learn? The Diffusion of Scientific Evidence on Generative AI [Dataset]". World Bank.
    Name
    World Bank Administrative Data
    Note
    Dataset names: `Position data.dta`, `Base Dataset.dta` These are restricted data sets, to require access please reach out to mshaukat@worldbank.org. `Base dataset.dta` contains anonymized division IDs and anonymized PMU IDs. `Position data.dta` contains binary indicators for managerial and administrative positions. Both datasets were sourced from World Bank HR, and kept secure and anonymized in line with the World Bank Group Personal Data Privacy Policy. Access date: 2025.
    Access policy
    Data access was granted directly to the study authors by the data owners/managers and it is not included in the reproducibility package.
    Citation
    World Bank. (n.d.). Administrative Data [Dataset]. World Bank.
    Data statement

    Data is forthcoming in the World Bank Microdata Library.

    Description

    Output
    Type
    Working Paper
    Title
    How Do Organizations Learn? The Diffusion Of Scientific Evidence On Generative Ai
    Description
    Policy Research Working Papers (PRWP)
    Authors
    Author Affiliation Email
    Mahvish Shaukat World Bank mshaukat@worldbank.org
    Andreas Stegmann University of Warwick andreas.stegmann@warwick.ac.uk
    Mattie Toma Warwick Business School mattie.toma@warwick.ac.uk
    Date of production

    2026-01-22

    Scope and coverage

    Geographic locations
    Location Code
    United States of America USA
    Keywords
    Information Diffusion Organizational Learning Evidence Use
    Topics
    ID Topic Parent topic ID Vocabulary Vocabulary URI
    D83 Search • Learning • Information and Knowledge • Communication • Belief • Unawareness D8 Journal of Economic Literature (JEL)
    D73 Bureaucracy • Administrative Processes in Public Organizations • Corruption D7 Journal of Economic Literature (JEL)
    M50 General M5 Journal of Economic Literature (JEL)
    C93 Field Experiments C9 Journal of Economic Literature (JEL)

    Disclaimer

    Disclaimer

    The materials in the reproducibility packages are distributed as they were prepared by the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this event do not necessarily reflect the views of the World Bank, the Executive Directors of the World Bank, or the governments they represent. The World Bank does not guarantee the accuracy of the materials included in the reproducibility package.

    Access and rights

    License
    Name URI
    Modified BSD3 https://opensource.org/license/bsd-3-clause/

    Contacts

    Contacts
    Name Affiliation Email
    Mahvish Shaukat World Bank mshaukat@worldbank.org
    Reproducibility WBG World Bank reproducibility@worldbank.org

    Information on metadata

    Producers
    Name Abbreviation Affiliation Role
    Reproducibility WBG DECDI World Bank - Development Impact Department Verification and preparation of metadata
    Date of Production

    2026-01-22
    Document version

    1
    Back to Catalog